Yesterday I (very, very belatedly) realized I could probably use the historical underground gold production data from the Cripple Creek Mining District as output data and regress this against the surface rock chip geochemical data provided in a mid-1960s USGS Professional Paper (625-A) to come up with a mineralization-predictive production function specific to Au-Te mineralized systems like those found at Cripple Creek.
It was relatively tricky to set up the regression leading to the production function whose results are displayed below because the Au-Te mineralizing environment contains two wholly different metal-precipitating processes in more or less the same volume of rock at more or less the same period of time. Sulfides are being deposited because of sulfate-reduction while tellurides are being deposited because of Eh change. Coming up with metal parameter proxies for these two different systems required some 'soul-searching'
In any case, now the algorithm has been produced, it's been used to transform/filter the CCAP stream sediment geochemical data referenced above with very gratifying results. See map legend comment, for example, concerning clean detection and classification of the very small Carbonate King Au-Te mineralization occurrence 21 miles west of Cripple Creek. Previously, that mineralization was transparent to my use of the CCAP geochem data.